Exploiting indigenous knowledge of subsistence farmers’ for the management and conservation of Enset (Ensete ventricosum (Welw.) Cheesman) (musaceae family) diversity on-farm

Background Enset (Ensete ventricosum (Welw.) Cheesman) belongs to the order sctaminae, the family musaceae. The Musaceae family is subdivided into the genera Musa and Ensete. Enset is an important staple crop for about 20 million people in the country. Recent publications on enset ethnobotany are insignificant when compared to the diverse ethnolingustic communities in the country. Hence, this paper try to identify and document wealth of indigenous knowledge associated with the distribution, diversity, and management of enset in the country. Methods The study was conducted in eight ethnic groups in the Southern Nations, Nationalities and Peoples’ Regional State. In order to identify and document wealth of indigenous knowledge, the data was collected mainly through individual interviews and direct on-farm participatory monitoring and observation with 320 farm households, key informant interviews. Relevant secondary data, literature and inter-personal data were collected from unpublished progress report from National Enset Research Project, elderly people and senior experts. Results Enset-based farming system is one of a major agricultural system in Ethiopia that serves as a backbone for at least ¼ of country’s population. Farmers used three morphological characters, two growth attributes, disease resistance and five use values traits in folk classification and characterization of enset. A total of 312 folk landraces have been identified. The number of landraces cultivated on individual farms ranged from one to twenty eight (mean of 8.08 ± 0.93). All ethnic groups in the study area use five use categories in order of importance: kocho yield and quality, bulla quality, amicho use, fiber quality and medicinal/ritual value. Of the 312 landraces 245 landraces having more than two use types. Management and maintenance of on-farm enset diversity is influenced by systematic propagation of the landraces, exchange of planting material and selective pressure. Conclusion It can be concluded that the existing farmers’ knowledge on naming, classification and diversity should be complemented with maintenance of the creative dynamics of traditional knowledge and transmission of the knowledge are crucial for constructing sustainable management.


Background
The Ethiopian highlands are a center of genetic diversity for enset, tef, sorghum, barley and finger millet [1]. Enset (Ensete ventricosum (Welw.) Cheesman) belongs to the order sctaminae, the family musaceae. The Musaceae family is subdivided into the genera Musa and Ensete [2]. Enset is an important staple crop for about 1/4 (20 million) of the population of the people living in the densely populated regions of South and Southwestern Ethiopia. The crop is grown in mixed subsistence farming systems, often in association with coffee, multipurpose trees, and annual food and fodder crops [3]. Enset is also used for livestock feed, fuel wood, construction materials, containers, and as a provider of shade to intercropped annual or perennial crops [4]. It is cultivated between 1500 and 3100 m above sea level (m.a.sl), where daily average minimum and maximum temperatures are 8 and 27°C, respectively [5].
The major food types obtained from enset are kocho, bulla and amicho. Kocho is fermented starch obtained from decorticated (scraped) leaf sheaths and grated corms. Bulla is obtained by squeezing out the liquid containing starch from scraped leaf sheathes and grated corm and allowing the resultant starch to concentrate into white powder. Amicho is boiled enset corm pieces, mainly obtained from young enset plants that are prepared and consumed in a similar manner to other root and tuber crops [6].
Studies indicate that numerous enset cultivars were identified in each region and the observed genetic diversity in cultivated enset in a particular area appears to be related to the extent of enset cultivation and the culture and distribution pattern of the different ethnic groups [7].
A clear understanding of the diversity and distribution of enset is important for crop improvement programs and for managing genetic resources. To measure the status of crop diversity in the field the most common method is counting named varieties. There are two main landrace diversity indices, namely: cultivar richness, which represents the number of landraces in a community, and cultivar evenness, representing the relative abundance of the individuals among the various landraces present in the community [8,9]. For farmers, genetic diversity means varietal diversity, which farmers can clearly distinguish on the basis of agro-morphological traits, phenological attributes, post-harvest characteristics, and differential adaptive performance under abiotic and biotic stresses [10].
Indigenous technical knowledge is the tool by which local people interact with the environment in order to meet needs and goals ranging from survival goals to that of achievement and esteem [11]. It is knowledge, which is unique to a local area, culture, or society, passed down from one generation to the next, usually through oral tradition. Indigenous knowledge has to do with theories, beliefs, practices, and technologies that local people have elaborated without any assistance from the modern, formal and scientific communities and/or institutions [12]. Indigenous people have a long tradition in maintaining biodiversity as a sustainable resource. Farmers have played and still continue to play a tremendous role in developing and nurturing crop genetic diversity. Many studies have shown that farmers in developing countries have intimate knowledge of environmental processes and make rational resource management decisions based on that knowledge [13].
The southern and southwestern part of Ethiopia has an extraordinary biological and cultural diversity. Recent publication on enset ethnobotany including those by [13,14] attempt to document farmers' indigenous knowledge on enset in some cultural groups at specific location. However, those documentations are insignificant when compared to the diverse ethnolingustic communities in the country. This paper seeks to contribute towards filling this knowledge gap, based on an empirical study of enset farmers in Ethiopia. The paper address the following main question: what are farmers' knowledge associated with the distribution, diversity, and management of enset in the country? The underlying assumption behind this question is that all farmers are equally likely to be knowledgeable about the crop.
Hence, the objectives of this study was to identify and document wealth of indigenous knowledge for folk naming, classification, distribution and abundance of enset landraces and understanding the corresponding knowledge related to utilization, management and conservation of enset landraces.

The study area
The SNNPR is one of the regions in Ethiopia. It is located in south and southwestern part Ethiopia, 4.43°-8. 58°N latitude and 34.88°-3914°E bordering Kenya to the south and South Sudan to the west and southwest, the Ethiopian region of Gambela to the northwest, and the Ethiopian region of Oromia to the north and east (Fig. 1). The region has a total area of 110,931.9 square kilometers lying within elevations of 378 to 4207 m above sea level [15]. The annual temperature is less than 10°C in the extreme highlands to over 27°C in the lowlands of the south. The regions are sub divided in to zones, which are organized in to weredas/districts. The zones are named based on the name of the dominant ethnic group for that specific location. The Regions are sub-divided into Zones, which are organized into weredas/districts. Within weredas, kebeles are the smallest administrative units.

Study site selections
The study was conducted in eight ethnic groups/zones (Dawro, Gedeo, Gurage, Hadiya, Kembata-Tembaro, Sidama, Silte, and Wolaita) in the Southern Nations, Nationalities and Peoples' Regional State (SNNPRS). The eight ethnic groups were selected for the following reasons: 1. The crop has coexisted with the people for centuries and enset production is pre-dominantly based on farmers' varieties. Hence, farmers' expected to have an established folk naming, classification system of appraisal of enset. 2. The ethnic groups had rich on-farm genetic resources of enset that made it suitable to study ethnobotanical descriptions [13,14] of enset. 3. In the region, enset cultivation is the center of the cropping system in which the entire farming system is based and the crop is the major food security and livelihood source [13,14].
Two wereda were selected from each ethnic groups based on enset diversity (Table 1). Then, two kebeles which are major enset growing areas were purposively selected from each wereda/district based on the importance of enset cultivation and information about enset distribution obtained from the Departments of Agriculture and Natural Resource of the respective zones.

Sampling
Multistage sampling technique was employed for selection of samples, zones, weredas and kebeles. All stages were selected purposefully from high (>2500 m.a.sl) and mid altitude (1500-2500 m.a.sl) [16] areas in consultation with stakeholders engaged in the subsector. Eight Zones, two weredas from each zone (16 wereda) and two Kebele Administration (KA) (Kebeles are the lowest administrative unit) from each wereda (32 KAs), were selected purposefully based on agro-ecology variant. A total of 320 households (40 household heads from each ethnic) over the selected ethnic groups in the two crop ecologies were directly monitored on farms. The survey focused on the investigation of farmers' folk knowledge for naming, classification, diversity and management of enset landraces in the region.

Data collection
Diverse data collection methods were employed in order to understand the many features for the acquirement of local knowledge of enset naming, classification, diversity and management in the center of diversity. The data collection was conducted mainly through: i) individual interviews and direct on-farm participatory monitoring and observation, ii) key informant and focus group discussions, and iii) secondary data and literature survey.

Individual interviews and direct on-farm participatory monitoring and observation
Before interviews were performed, informal conversation was conducted with 20 inhabitants of the enset community with the objective of determining which type of information needed to be collected. Based on these conversations, semi-structured interviews were designed and data collected with the head of the household or the person responsible for maintenance of the enset plantation. Three hundred twenty farmers were interviewed A B Fig. 1 a Detail zones map of the study region. b Map and Position of the study region in Ethiopia and directly monitored on farms, over the selected weredas in order to assess the farmers' ethnobotany knowledge on enset.
The questionnaire covered different topics such as information about the study area, landholdings, crops commonly grown and specific information on the use and management of enset. The detailed information was focused on enset diversity, cultural practices, source of planting materials, and traditional use values of enset. The respondents were also asked about their perception on enset production constraints and their indigenous knowledge about the disease.

Key informant interviews
In order to assess the general indigenous knowledge of farmers' in each ethnic group: key informants up to five per KA, community leaders, local administrations, and MOA (Ministry of Agriculture), and other members in each ethnic site were interviewed.

Secondary data and literature survey
National Enset Research Project progress report was visited for secondary data and personal communication and discussion with elderly people and senior experts in line with ethnobotany tradition of enset. Literatures on enset culture were reviewed from published and unpublished sources and reports.

Data analysis
Informal discussion with elderly farmers, and key informants were carried out to validate the information gathered from individual interviews. Lists of all landraces described throughout the study area were summarized after grouping known synonyms or names that refer to the same landraces in each wereda with the help of elderly farmers.
Collected survey data were subjected to descriptive statistics (frequencies, percentages, and average) using SPSS Ver. 16. Landrace richness, diversity and dominance per farm were calculated using Microsoft excel 2010. Richness was calculated as the total number of landraces per farm and averaged this figure per ethnic group. Abundance was calculated as the total number of individual plants of each landraces per farm/household. Frequency was estimated as the number of individuals of a landraces with respect to the total number of landraces composing the enset farm. With these parameters we calculated the ecological importance index of each cultivar per farm.
The Shannon and Weaver [17] and Simpson [18] diversity indices are two of the most widely used measures of heterogeneity [19]. Both of them were calculated for all the surveyed zones. The Shannon-Weaver diversity index accounts for both abundance and evenness of the landraces present and can be increased either by greater evenness or more unique landraces. It was calculated using the formula, H' = − Σ pi ln pi, [19]. Where pi, the proportional abundance of the i th landrace. Then we calculated the dominance as a measure representativeness of each landrace through the Simpson index. Simpson's Index of Diversity (1 -D) was computed for all the zones and all the landraces using the function: Simpson's Index of Diversity (1-D) = 1-∑ (n/N) 2 .
where, n i = the frequency of the i th landrace, frequency being the number of farms in which the landrace is found in the district, and N = the total number of farms surveyed in the zone. Equity, the proportion of the observed diversity with respect the maximum diversity expected was calculated through the Pielon index: J = H'/H'max, in which J is equity; H' = diversity; H'max = maximum diversity, H'max was calculated as the ln(S) S being the number of landraces in a sample. Pearson's correlation coefficient was used to compare diversity and distribution values at different ethnic groups.
We used a multiple use curve [20] concept to describe the rate at which ethnobotanical data is collected, check whether the essential part of the available information on the landraces had been collected. This curve plotted the cumulative number of uses recorded against the number of informants. To analyze the use values of the landraces, we regrouped the uses into broad categories, where each category contained uses of a similar nature. In this way, three main categories were created, namely; food (kocho yield and quality, bulla quality, amicho use), fiber (fiber quality) and medicinal/ritual categories. Food and medicinal categories refer to use by both humans and animals.

Strategic importance of enset
Enset-based farming system is one of a major agricultural system in Ethiopia that serves as a backbone for at least one-fifth of country's population. Enset has been selected as a typical multipurpose crop of which every part is thoroughly used for food, feed, medicinal, construction and ornamental purposes. Throughout the growth stage the corm, pseudostem and leaves are sued for various purposes. Enset is intimately associated with the daily lives of the farmers. Owing to these facts, farmers indicated that, 'enset is everything for us'. 'It is our food' (Fig. 2a), 'it is our plate' (Fig. 2b), 'it is our house' (Fig. 2c), 'it is our bed' (Fig. 2d),'it is our bag' (Fig. 2e) 'it is our cattle feed' (Fig. 2f) and it is our medicine (Fig. 2g). It is the most important crop in the farmers' livelihoods and security.

Indigenous knowledge in naming and classification
Ensete is the genus name, while different ethnic group use different vernacular terms as a local name for Ensete. In the study area, Ensete ventricosum is identified  (Table 2). Farmers in the study area use a combination of similar criteria to name and classify enset landraces (Table 3). They classify their landraces and give different names based on several attributes that distinguish these landraces from one another. Three morphological characters (midrib color, petiole color, and leaf color), Growth attributes (vigor, maturity), disease resistance and use value food (kocho yield and quality, bulla quality, amicho use), fiber quality and medicinal value were the major criteria used by farmers. The interviewees referred first to the morphological characters (48 %) (Fig. 3) of any enset landrace when asked for key classifying characteristics. The food usage, food quality, and other use value characters were usually mentioned as those of second importance for classification. It is witnessed that the names given by all enset growing farmers to the different landraces and the classification criteria are generally consistent.

Level of on-farm richness, diversity and pattern of use
We recorded a total of 440 folk varieties (landraces) across the eight ethnic groups. From the total 128 (29 %) landraces shared the same name in at least two ethnics and the total number of landraces reduced to 312 (Table 14). As farmers over years have selected their landraces for multipurpose values, they do group them according to the use values landraces renders. Each landrace is clearly distinguished by its vernacular name and peculiar characteristics. Of the 312 landraces 288 were reported to be known by all of the interviewees, whereas the 24 landraces were found in less than 5 % of the respondents' farm. Based on the total number of different landraces recorded (richness of the ethnic group) and the number of enset landraces per farm, Dawro farmers' had the highest number of landraces (75) accounting for 24 % of the total number of recorded landraces across the study area. In contrast, the lowest richness was found in Gedeo farmers' with 26 landraces accounting for 8.33 % of the total number of recorded landraces ( Table 4). The number of landraces cultivated on individual farms  Kocho yield Low (<9.9 t ha −1 yr −1 ) 9 Medium (9.9 to 20 t ha −1 yr −1 ) 53 High (>20 t ha −1 yr −1 ) 3 8 Bulla ranged from one to twenty eight (mean of 8.08 ± 0.93) ( Table 4). Average number of landraces per farm ranged between 10.43 for Silte to 3.55 for Wolaita. Dawro and Sidama with 10.2 and Gurage with 9.45 landraces per farm had high farm level richness (Table 4). Diversity indices for the eight ethnic groups studied were computed from the numbers of landraces present on the 40 farms within the ethnic (Table 4). Although ethnics differed in richness, they were similar in diversity. The Simpson's 1-D ranged between 0.97 (Dawro) to 0.9 (Gedeo), H′ ranged between 3.71 for Dawro to 2.6 for Gedeo, while evenness also had a very narrow range: 0.89 for Gurage to 0.8 for Gedeo (Table 4). Both the H' and 1-D indices were highly correlated with landrace number at each ethnic (r = 0.90 and 0.70). All these values indicate the high enset diversity in these eight ethnic groups.
All ethnic groups in the study area use a combination of different criteria to group enset landraces. We recorded three use categories, as defined by (25), in order of importance: Food (kocho yield and quality, bulla quality, amicho use), fiber (fiber quality) and medicinal/ritual value as described in Table 3. Of the 312 landraces: only 11 landraces having one use type, 56 landraces having two use types and a total of 245 landraces having more than two use types (Fig. 4). In addition, Fig. 5 shows the comparative result of the use categories according to the ethnic groups. Fair analysis between ethnic groups revealed that the highest value for food (kocho yield and quality) were (≥35 house hold/ethnic) observed in all ethnic groups.
Almost all of the landraces used for good kocho and bulla yield and quality have got a wider distribution and diversity ( Table 5). The fiber uses showed higher values for all ethnic groups. Farmers also reported enset landraces having longer and/or stronger fibers, and higher fiber yield and quality (Table 6). Forty two landraces were identified by farmers for amicho use value (Table 7). In addition, some enset landraces are known by farmers to have medicinal value for both humans and animals. These landraces are poorly producing and to be maintained for special traditional or religious uses (Table 8). Almost all landraces in this category have got sweet amicho test therefore both categories share more than 50 % of the landraces. In addition to the above use value; farmers in each ethnic group use biotic and abiotic tolerance as a trait for diversity maintenance. Fifty and thirty three landraces were identified by farmers as tolerant to enset bacterial wilt and drought (Tables 9 & 10).

Indigenous knowledge on the management of enset diversity
People in the study area maintain their enset farm with considerable structured planting, diversity and flexibility that support production of this livelihood crop. They have managed to select landraces that adapt the local environment and that give multiple benefits. According  to the information we obtained during individual interview, key informant and focus group discussion, management and maintenance of on-farm enset diversity is influenced by: (i) systematic propagation of the landraces, (ii) exchange of planting material (iii) selective pressure.
(i) Systematic propagation of the landraces Systemic propagation of the landraces is practices used by all farmers in the study area to adjust and to maintain the landrace diversity. Almost all farmers in the study area use corms of 3 to 4 years old enset plants with some portion of the pseudostem to produce enset seedlings ( Fig. 6 & Table 11). Almost all respondents indicated that there are three to four growth stages or frequency of transplanting before harvesting (Table 12). The informants indicated that the propagation starts from the third stages of transplanting (Fig. 6e). Farmers traditionally practiced removal of the central shoot and removal of the apical dominance corms ready for burring (Fig. 6a). Hypothetical question posed in the interviews was what happen if you plant the corm without removal of the central part? The respondents indicated that the removal of the central area helps the propagated corm to produce more number of suckers (≥50 suckers /corm) for next season multiplication (Fig. 6b). The first sucker production stage stays 1 year after emergence from the buried corm (Fig. 6c). In the second stage, the produced multiple suckers from the buried mother corm detached and planted in rows with two to three suckers in a group, or in rows of single plants (Fig. 6d). A consecutive transplanting produces the third stage (Fig. 6e). Farmers' indicated that the third stage is used as both the source of mother corm for sucker multiplication and harvested for consumption when there is less amount of food in the stock. At the end of the third stage, the suckers are transplanted a fourth time to the permanent field (Fig. 6f ). The total time required from first planting to harvesting can be around 7-8 years. The propagation usually carried out in the dry season (November to early February). Farmers propagate a diverse landraces available in the farm. Some multipurpose landraces are propagated by the majority of households interviewed. (ii)Exchange of planting material Traditional planting material exchange system is an important source of diversity for majority of farmers. Out of the 320 farmers interviewed 249 farmers use corms from their own farms (Fig. 7). One fourth of the 320 farmers' interviewees mentioned that they often hand out or sell corms/planting material to neighbors or fellow villagers. Neighbors, relatives, and market were the sources of planting material and exchange, gift, purchase and free distribution

Strategic importance of enset
Enset is well-established, sustainable, and environmentally resilient farming system that contributes to food security of farmers and, in particular it serves as food security crop in densely populated areas. Enset needs to be present in farmers' pits throughout the year. Enset is the most important crop in the region. According to 2011 CSA [16] report 3,020,143 km 2 of land is covered by enset crop and about 6.9 million quintals of enset yields were produced in 2010/11 production season. All farmers are using the landraces developed by the community [21]. These landraces have been grown onfarm thousands of years. These enset-growing traditions still continue in the current generation. Enset represents an important cultural plant in the region. This

Indigenous knowledge in naming and classification
Farmers' rich knowledge that is accumulated on the crop over many years has played a significant role in naming, characterization and maintenance of the existing genetic diversity. Enset producing farmers have their own folk naming and classification system to distinguish one landrace from the other. Sometimes it is difficult to understand and reclassify, even while watching them to characterize. The classification of enset landraces has been accommodated by phenotypic differences, unique traits and specific uses of landraces. As pointed out by [4,13,14], these are common characteristics of folk classification systems in enset. Folk nomenclature is an integral part of the variety management in enset farming systems [24,25]. In view of this, the multitude of names in various folk taxonomic levels indicated the occurrence of on farm genetic diversity at infra-specific level. As indicated by [13], landrace names given by farmers' have been used as farmers' diversity unit for estimating unit for the extent and distribution of enset diversity as well as ex situ collection. This is also in agreement where folk taxonomy is used to highlight the amount of genetic diversity [18,26,27]. In this study, over 300 landrace names (Table 14) have been identified which indicated the level of on farm genetic diversity. The meaning of the names of most landraces is not known. It is difficult to know unless the people who named it or the place of origin are traced back. It has been repeatedly reported that unexplained meanings of folk names were common in other ethnic groups [13]. A similar pattern was observed in other crops like sorghum and rice [28,29]. Enset landraces were commonly exchanged and distributed according to the folk names. The finding of this study (Table 5) and other similar studies [13,14,21]  depict identically named landraces were also reported in more than one ethno-linguistic communities. Folk classification can help in identifying the comparative value of landraces (for example Tables 6,7 & 8) for proper characterization and pre-breeding activities. A similar study on sorghum in Ethiopia [28] and rice in Nepal [29] has shown that name of the varieties indirectly related showed the functional value for the variety. Commonly, knowing folk names and classification may distinguish varieties that are actually genetically very closes. Farmer's in one household generally knows which households certainly have named varieties and their particular agronomic and use value related characteristics. Knowing folk taxonomy also helps in developing planting material distribution, flow channels, and regional landrace map. Thus, even if landrace names and classification are a necessary basis, they are not sufficient to describe genetic diversity. Integrative indicators have been designed e.g., complementing the naming and folk classification with parameters of genetic diversity.  Fig. 6 Systematic propagation of enset. a mother corm ready for burring; b suckers emerged from the mother corm; c 1st stage transplanting; d 2nd stage transplanting; e 3rd stage transplanting; f Matured enset ready for harvesting Our data thus needs to be complemented by phenotypic and genotypic information which helps to avoid redundancies and optimizing the efficient conservation and sustainable use of the crop.
Level of on-farm richness, diversity and pattern of use Enset farming systems are rich in landraces diversity. In the study area we recorded a relatively high landraces (312) richness of enset. For instant, in previous studies, comparable results were reported by [21], who described 218 different enset landraces from seven ethnic groups. One hundred eleven enset landraces were also reported from nine growing areas of Ethiopia [7], while [13] described 67 enset landraces from Wolaita zone of the southern region. The number of enset landraces in this study is far higher than what was reported by previous studies which were conducted in zones with similar climatic and altitudinal factors. For instance, [21] reported the presence of 41 landraces in Dawro, which is far below the number of enset landraces reported in the present study. During discussion with the farmers it has been observed that, there were more than 100 enset landraces grown in each locality a few years back, however, farmers reported that most of the landraces were lost due to EXW. Tesfaye [24] also found out that in Sidama zone farmers reported names of 20 enset landraces which were not encountered in any of the farms that were visited. Some enset landraces might have been totally lost from farmers' fields. Enset is a multipurpose crop which is utilized for different use values. Based on their use value and folk classification large differences were evident between landrace abundance and distribution in the region. Some landraces, particularly those having merits of better kocho yield and quality have got a wider distribution within and between ethnic groups/zones. For example, the enset landraces 'Shododenia' and 'Addo' were encountered on respectively 37 and 34 (92.5 and 85 %) farms visited in Dawro and Sidama, but were not found in any other surveyed zones. Some landraces had a very high local abundance at one or two locations and were absent from the rest. For example Shodedenia was encountered on 100 % of the farms visited in Dawro. It was encountered on all the 40 (100 %) farms visited in Dawro. Likewise, [24] reported a small number of landraces (for instant Genticha) playing a dominant role in Sidama zone. Our study revealed that the highest use values of the landraces were found in the region which also corresponds to where the landraces had the highest abundance in the farming system. This suggests a positive relationship between plant abundance and use. These findings corroborate the "apparency hypothesis" which describes dominant, large and more abundant plant species as having the highest use values.
Enset bacterial wilt, caused by Xanthomonas campestris pv. musacearum, is the most important biotic constraint to enset cultivation [6]. In order to alleviate this biotic stress farmers integrate EXW tolerant landraces in their farms. The kocho yield of these disease tolerant landraces is however below average [26,27]. Moreover, some enset landraces are known by farmers to have medicinal value for both humans and animals. These landraces are most often poor yielding and are only maintained for special traditional or religious purposes/ uses. Those landraces are reported to heal bone fractures, are used for treating diarrhea and during child delivery i.e., assisting the discharge of the placenta. Most reports of medicinal and ritual uses of enset indicate that farmers' intentionally maintain the landraces together with other landraces. For example, [27] described 14 enset landraces based on their medicinal and ritual use value. Likewise, [26] reported a number of different enset landraces to have medicinal and religious (ritual) significance for preventive treatment, healing and other therapeutic purposes and as protection against evil spirits. Farmers also categorize enset landraces as male or female based on different characteristics [21,30,31]. However, the designation of landraces as 'male' or 'female' is not linked to their reproductive biology. According to farmers, the male enset landraces are drought tolerant. This designation is very important for maintaining landraces for amicho use value. Female landraces are described by farmers as less vigorous, susceptible to disease, having a higher kocho quality and producing edible and tasty amicho [31]. In addition, they are early   [14,25] revealed that in most ethnical groups farmers maintain a single landrace for multiple uses. In some cases, poorly producing landraces continue to be maintained for special traditional (e.g., medicinal value) or religious uses. Farmers often maintain low yielding landraces that have medicinal values [25]. Similar observations have been made in bananabased communities in Uganda [32] or in rise systems in Asia [33]. Knowledge of the local usage of enset resources is essential for the elaboration of conservation strategies. This is the first time that the use values according to various ethnic groups in the study area have been evaluated in detail for enset. Overall, we found less diverse ethnic variation in knowledge and use values of enset, as has been found for difference within the same ethnic group [13,14]. In general, this study and the previous studies have shown that different ethnic groups in the enset farming system demonstrated the existence of considerable amount of indigenous ethnobotany knowledge. High landrace diversity in a region may indicate extended periods of enset cultivation and a more subsistence form of production.

Indigenous knowledge on the management of enset diversity
In the region, farmers' manage local enset landraces within traditional production and processing systems oriented towards meeting household subsistence needs. Both women and men as producers, selectors, processors and marketers of enset are traditionally the custodians of in situ conservation. Farmers generally choose planting material from their existing mats. Farmers plant their enset landraces mixed on their fields, usually two or ten, but sometimes up to 20 landraces in one plot. It is traditional to use a corm and sucker as planting material and use of different transplanting stages in enset producing farmers. It was found that many households could propagate enset landraces in at least two ways and this flexibility of propagation might also reflect a relative preference for growing in a large area. A similar observation was also reported in other enset growing areas [13,30,31]. However it is yet to be identified whether such variations in propagation have some implications on maintenance of diversity in situ. Farmers observe and select the landraces based on their planting intentions for the coming year than the proportion to the quantity they have. This scenario has been maintained by the systematic propagation of 3-4 years old enset landraces. Other study [13] revealed that regular propagation and     harvesting restrain; organized assemblage and arrangement of landraces in the home gardens and landrace composition regulation in the home gardens have been the major factor for indigenous management and maintenance of enset landraces on-farm. The rich selection experience on indigenous crop such as enset is also applied to other crops like sorghum [24]. The number of landraces grown at a given locality, their genetic similarity and the areas they occupy over time and space are influenced by planting material source, exchange and supply. Most planting material exchange is local, though a small proportion extends beyond the local group of villages reflecting relationships among neighbors and kin in most cases. All landraces used in the region are local farmer-named varieties. Among the surveyed farms, most farmers produce their own planting material. In addition farmers in the region have fixed systems to ensure the sustenance of planting material supply for each season. Farmers in cereal based farming system have well-established systems to ensure self-sustaining seed supply system and they often operate the exchange of planting material in the local market [34]. In general, on-farm conservation enhances continued source and supply of genetic material and continued diversity-based agriculture as compared to monoculture by ensuring intraspecific and interspecific diversity of crops. Farmers themselves perceived an advantage in continuing to grow diverse traditional crops and their participation in conservation of a traditional seed system proved to be self-sustaining.
Similarly farmers in the region quite frequently practices grow their landraces in mixture to stabilize their crop production, especially under adverse growing condition. Farmers may retain their preferred landraces over many years, often claiming they received no external inputs of seed/planting material. Plant diseases can also reduce the level of biodiversity or limit the variety of plants grown in an area. It have been observed that, the genetic base has been vulnerable to a range of very damaging biotic and abiotic stresses such as Enset Xanthomonas wilt (EXW), enset root mealy bug, leaf hopper, mole rat, Porcupine, wild pigs, corm rot, and drought. It is the EXW which has had the greatest impact on enset production. In Hadiya zone Lemu wereda 30 % of enset crop affected by EXW [35]. Therefore, farmers are forced to develop their copping strategies. Almost all surveyed farmers in the region practice cropping and dietary patterns change and grow more number of disease resistant plants as a strategy for the management of the disease. For instance, [36] indicated genetic diversity can be seen as a defense against problems caused by genetic vulnerability. To reduce the likelihood of spread, establishment and growth of EXW in enset crops, a systematic operational approach to the management of EXW should be adopted. This should include giving training to farmers on appropriate production practices, using healthy suckers and planting in clean soils. Future efforts surely need to focus on developing core collections representative of the widest possible genetic diversity for enset improvement and using this to strengthen in situ or on farm conservation.

Conclusion
The information collected in the region and presented here shows that a certain wealth and diversity of knowledge regarding traditional naming, uses of plants and diversity management as a part of the cultural heritage of the community. Farmers' have been growing enset for many years. The farmers' knowledge and enset have been coevolving together. This has resulted in the prevalence of rich indigenous knowledge of the farmers. Any attempt to improve the crop needs to take in to account the farmers knowledge and experience.
Folk naming and classification are not consistent across all ethnic groups. The inconsistency is highly related with the ethnolinguistic variation in the region. Integrated folk-formal classification and characterization will be imperative for management and utilization of on farm genetic resources.
Our study confirms that the landrace diversity and distribution makes it possible to gain a general picture of the uses made of such crop on a macro-scale. A principal conclusion from the present study is that the biggest uses of landraces, in terms of the number of citations in the literatures, are for kocho, bulla, amicho, fiber and medicine. Certain traditional practices (for example spiritual or rituals) also lead farmers to maintain small quantities of uncommon landraces that may not produce well. This scenario points to the importance of use value based and other criteria similarity and differences for landrace diversity maintenance and management. Hence, formal enset improvement program needs to positioned in to multipurpose enset variety development scheme and include farmers and their knowledge in the researchextension continuum. Landrace diversity in the region is affected by a number of factors. EXW is the main factor limiting enset richness and diversity. Any attempt to improve enset has to give emphasis on enhancement of farmers' varieties and a systematic operational approach to the management of EXW.
It can be concluded that the existing farmers' knowledge on naming, classification and diversity should be complemented with maintenance of the creative dynamics of traditional knowledge and transmission of the knowledge are crucial for constructing sustainable management.